hmm.. so I have to keep mining with Eligius in order to get back to expected return. I'm not particularly well versed in the mathematics of probability.. but isn't this a bit like the gambler's fallacy? ie just because I've tossed a run of 10 tails in a row doesn't mean I'm more likely to get a run of heads. Does the maths really say It'll come good for me?
It is possible that you have bad luck "forever" according to gambler's fallacy. Be aware that arsbitcoin had this issue with their payout scheme and towards the end all the miners were out a LOT of btc and the luck slowly got better but never recovered the massive negative state they had reached and then the pool shutdown and the miners never got it back.
how can you call -1.138% Massive negative state...wondering what adjective you will use for itzod's -9.611%...
Heh that's just plain awesome and I wasn't aware of their magnitude of fail. Nonetheless, it meant people were a long way behind in payments on ars. I only bring this up because said payment scheme is relevant to this pool (which I don't mine at...)
The problem with luck and poolwide hashrate is that generally when a pool is going through badluck swings alot of users jump ship and then once they do hit some good luck recovery spree there is either fewer GH available for the duration or fewer users who left then returns.
If you aggregate this process you end up with arsbitcoin as the result and that is the single reason I just dont think SMPPS is viable at all longterm unless you have a way to keep your starting avg hashrate longterm the same or higher. Any GH drop and usually its substantial during badluck period in effect slowly kills the negative buffer overtime with little hope to recover as we saw with arsbitcoin.